In vitro diagnostic tests to identify and treat diseases have become common tools in hospitals, homes and physician's offices. Biological fluids such as blood, urine or cerebrospinal fluids, which may at times contain blood, are the most frequently employed biological samples its for such tests.
Blood contains many different components, some of which are present in strikingly varied concentrations from sample to sample. The percentages of both red and white blood cells in whole blood, for example, can vary among normal individuals, and even in the same individual over time, and in particular under pathological conditions. This large variation coupled with other factors such as storage conditions, coagulation, and the fragility of red blood cells, produces considerable technical problems in performing diagnostics using blood-containing samples.
Whole blood is usually separated into various fractions prior to testing. Among other things, separation into fractions can advantageously compensate for differences in hematocrit values, and in other ways reduce potential interference in up stream or down stream biochemical assays. Frequently employed fractions are serum, plasma, white cells, red blood cells and platelets. The term "plasma" is used herein to mean any fluid derived from whole blood from which a substantial portion of the cellular components has been removed.
Blood separation technologies can be conceptually grouped into three categories--centrifugation, filtration, and separation.
Centrifugation
Blood separation is routinely achieved by centrifugation. Centrifugation is generally desirable because: (1) centrifugation can generally separate cellular components from serum or plasma at an efficiency of greater than 95%; (2) centrifuges do not require highly trained personnel to operate; and (3) centrifugation allows concurrent processing of multiple samples in under 15 minutes. Centrifugation of blood is, however, also problematic. For example, centrifuges are expensive, involve multiple steps, are often unavailable at points of care such as bedside, schools or at home, and usually require electrical power for operation. Centrifugation also generally requires at least several milliliters of blood.
Filtration
Many filtration techniques are known for separating various components from blood. U.S. Pat. No. 4,987,085 to Allen et al., for example, describes a filtering system with descending pore size using a combination of glass fiber membranes and cellulose membranes. U.S. Pat. No. 4,753,776 to Hillman et al. discloses a glass microfiber filter using capillary force to retard the flow of cells. U.S. Pat. No. 4,256,693 to Kondo et al. discloses a multilayered chemical analysis element with filter layers made from at least one component selected from paper, nonwoven fabric, sheet-like filter material composed of powders or fibers such as man-made fibers or glass fibers. U.S. Pat. Nos. 3,663,374 and 4,246,693 disclose membrane filters for separating plasma from whole blood and U.S. Pat. Nos. 3,092,465, 3,630,957, 3,663,374, 4,246,693, 4,246,107, 2,330,410 disclose further filtration systems, some of which make use of small-pore membranes.
Known filtration techniques generally reduce the volume of blood required to only a few drops. Many filtration tests therefore contemplate using only about 25 to 75 .mu.l of whole blood. Some filtration techniques have even been developed that require only about 5 to 50 .mu.l of whole blood. In most applications, filtration occurs directly on a test-strip in which the filtration surface is placed above the reaction zone or zones of the strip. Filtration in these formats also reduces or eliminates the availability problems associated with centrifuges.
But these advances often create entirely new problems. For example, filters tend to retain significant amounts of plasma, and analytes present in low concentrations are frequently difficult to detect in the serum derived from small volumes of blood. Existing filters also tend to clog, and have undesirably slow flow rates. Agglutinating agents are often mixed with whole blood to reduce clogging and to improve flow rates, (see U.S. Pat. Nos. 5,262,067, 5,766,552, 5,660,798 and 5,652,148), but these problems remain.
Efforts have been made to improve the flow rate by modifying the force employed against the filter. But choices here are fairly limited. Filters are relatively simple to produce and use, but tend to cause excessive hemolysis of red blood cells. Capillary action, a phenomenon in which water or liquid will rise above normal liquid level as a result of attraction of molecules in liquid for each other and for the walls of a capillary can also be used. Capillary action, however, is generally too weak to effect rapid separation of large volumes. (See, for example, U.S. Pat. Nos. 5,660,798, 5,652,148 and 5,262,067). Moreover, separation of plasma by capillary action tends to retain a relatively large amount of fluid within the wicking membrane, or a collection membrane. This in turn may necessitate testing the wicking membrane or the collection membrane or both, or eluting the retained material from the membranes.
Solid-Phase Separation Solid-phase separation typically involves a surface having binding to a target, the surface acting to immobilize and remove the target from a sample. Exemplary solid-phase separation techniques are binding chromatography, binding separation using beads, and hollow fibers separations.
One particularly advantageous type of solid-phase separation is magnetic separation, in which a target is captured by magnetically attractable (paramagnetic) beads. Since no physical barriers are present, as would be the case with filtration separation, magnetic separation tends to be relatively gentle. In U.S. Pat. No. 5,514,340 to Lansdorp and U.S. Pat. No. 5,123,901 to Carew, for example, magnetic wires are employed in batch processes to separate magnetic particles from a fluid. In U.S. Pat. No. 4,663,029 to Kelland et al. and U.S. Pat. No. 5,795,470 to Wang, magnetic particles are separated out from a fluid in a continuous flow process. Still other methods published for example in U.S. Pat. No. 5,536,475 to Moubayed, employ rocking separation chambers and multiple magnets to separate magnetic particles from a fluid.
One of the major limitations of applying known magnetic separation to blood separation is that multiple anti-ligands are required to remove all of the various types of cells and sub-cellular particles. Red blood cells, lymphocytes, monocytes, and platelets, for example, have different surface antigens, and do not specifically bind to any one antibody. Furthermore, lack or absence of ligands on the cells due to pathological conditions, genetic diseases or genetic variations or life cycle of cells generally reduce the efficiency with which the anti-ligands bind with the target cells.
The problems with known magnetic separation devices are exacerbated with increasing sample volumes, especially sample volumes over one milliliter. Since many diagnostic applications require serum volumes of up to one milliliter to satisfy the requirements of multiple tests or batteries of tests, magnetic separation has not been particularly useful. Moreover, assays such as glucose or hemoglobin tests are highly susceptible to interference caused by biological or chemical substances in the sample, including proteins, bilirubin, and drugs.
Thus, there is still a need to provide improved methods and apparatus for separating blood into its constituent parts, and especially for separating plasma from whole blood.